Spatial and environmental effects on a rock‐pool metacommunity depend on landscape setting and dispersal mode

Empirical studies on structuring mechanisms of metacommunities usually focus on the major roles of environmental filtering and dispersal. Recent works suggest that the relative importance of these structuring mechanisms differs among organisms with different body size, taxonomic affiliation, and dispersal abilities, and also depends on spatial extent and environmental heterogeneity. However, the effects of physical connectivity among sites and dispersal mode are less commonly considered explicitly in field metacommunity studies. We analysed a rock‐pool animal metacommunity, comparing both environmental and spatial effects between a set of pools in a ravine setting, with ephemeral connecting waterways, and another in a hill setting, without such connections. We also analysed the relative role of structuring mechanisms influencing active versus passive dispersers in the metacommunity. We used permutational multivariate analysis of variance and analysis of the multivariate homogeneity of group dispersions to compare environmental and species variation between landscape settings. Variation partitioning was applied to determine the percentage of species variation explained by environmental and spatial variables. The relative influence of the structuring mechanisms depended on both the landscape connectivity context and species dispersal mode. Species sorting drove active‐disperser metacommunities in both isolated and waterway connected pools, suggesting that these animals had a dispersal rate among the environmentally suitable sites that was adequate to compensate for extinctions at the spatial scale we considered. In addition, beta diversity of active dispersers and species‐sorting effects were higher in the set of ravine rock pools that were more environmentally heterogeneous and connected by waterways. In contrast, species sorting structured passive‐disperser metacommunities only in the set of pools with connecting waterways, whereas spatial constraints more strongly affected passive dispersers in the relatively more isolated hillside pools. Overall, environmental variables had a greater effect than spatial variables on rock‐pool metacommunities at the scale we studied, with the exception that passive dispersers in rock pools unconnected by waterways did tend to have spatially constrained distributions.